Handle colander.null for mapped schemas in objectify

[tsauerwein]

I am having a schema like the following:

    class DocumentGeometry(Base):
        __tablename__ = 'document_geometries'
        id = Column(Integer, primary_key=True)
        document_id = Column(
            Integer, ForeignKey('documents.document_id'),
            nullable=False)
        geom = Column(String)

    class Document(Base):
        __tablename__ = 'documents'

        document_id = Column(Integer, primary_key=True)
        geometry = relationship(DocumentGeometry, uselist=False)

    schema = SQLAlchemySchemaNode(Document)

When I try to use schema.objectify({geometry: colander.null, document_id: null}), I was seeing the following error message:

    value = self[attr].objectify(dict_[attr])
  File ".../ColanderAlchemy/colanderalchemy/schema.py", line 687, in objectify
    for attr in dict_:
TypeError: '_null' object is not iterable

When the value for attribute geometry is colander.null, None should simply be used as value instead of trying to call objectify with it.